1. Field
The present invention relates to a network relay apparatus that has a redundant structure with a primary system and a secondary system (i.e., working system and a spare system), and relays data between different networks in a composite network.
2. Description of the Related Art
A technology related to redundancy that provides a spare system in addition to a working system in a network, in case a failure occurs in a node or in a dedicated line that connects nodes has been developed (see Japanese Patent Application Laid-open No. H11-55285 and Japanese Patent Application Laid-open No. 2005-517869).
In recent years, a composite network is configured with Ethernet (registered trademark, hereinafter omitted) networks connected to a synchronous optical network/synchronous digital hierarchy (SONET/SDH) network. In such a composite network, it is also important to handle the similar failures.
The composite network includes an optical transmission apparatus (hereinafter, “network relay apparatus”), which is a node that links the Ethernet network and the SONET/SDH network.
Conventionally, the network relay apparatus in general is configured as described in FIG. 9 to secure redundancy in the node and thereby to prepare for the failure.
The redundancy in the network relay apparatus will now be explained in detail below with reference to FIG. 9. A network relay apparatus 10 includes a switch fabric 20 that performs transmission/reception processing of data with respect to the SONET/SDH network, and an Ethernet interface unit 40 that performs the transmission/reception processing of data with respect to the Ethernet network. The network relay apparatus 10 also includes a switch fabric 30 and an Ethernet interface unit 50, as the respective spare systems for the switch fabric 20 and the Ethernet interface unit 40.
A data relay unit 60 that transmits and receives data to/from the network relay apparatus 10 at a side of the Ethernet network includes a port 70a, a port 70b, a port 80a, and a port 80b. The port 70a and the port 80a are respectively connected to the Ethernet interface unit 40 and the Ethernet interface unit 50. In the data relay unit 60, a communication path for data is established between the port 70a and the port 70b, and between the port 80a and the port 80b. 
Referring back to the network relay apparatus 10, the switch fabrics 20 and 30 replicate the received data and output it to the Ethernet interface unit 40 and the Ethernet interface unit 50.
The Ethernet interface units 40 and 50 include switches 41 and 51, respectively, and selectively receive data from the switch fabric 20 or the switch fabric 30. The network relay apparatus 10 breaks the link between the Ethernet interface unit 50 and the port 80a of the spare system, so as not to receive data twice from the data relay unit 60.
In this manner, a communication path for data as shown by a solid line in FIG. 9 is established. If a failure occurs in the switch fabric 20 while main data is actually relayed as shown by solid line arrows (at this time, the switch fabric 30 is also receiving the same data as shown by dotted line arrows), for example, the network relay apparatus 10 handles the failure by switching the switch 41 and relaying the received data by the switch fabric 30.
Similarly, if a failure occurs in the Ethernet interface unit 40 while the main data is actually relayed as shown by the solid line arrows, the network relay apparatus 10 handles the failure by connecting the link between the port 80a and the Ethernet interface unit 50 and relaying the received data by the switch fabric 30. Further to the link connection, the network relay apparatus 10 can handle the failure by switching the switch 51 to relay the data received by the switch fabric 20, if there is no failure in the switch fabric 20.
To obtain redundancy in the node with the configuration of the network relay apparatus 10 as described above, it is also necessary to install the port 80a and the port 80b of the spare system in the data relay unit 60, corresponding to the communication path of the Ethernet interface unit 50 of the spare system in the network relay apparatus 10.
The network relay apparatus 10 breaks the link between the Ethernet interface unit 50 and the port 80a, and prepares for a failure in the network relay apparatus 10, until a failure occurs in the Ethernet interface unit 40 of a working system.
The ports 80a and 80b of the date relay unit 60 well serve the purpose of preparing for the failure in the network relay apparatus 10. However, if the failure does not occur, the ports 80a and 80b will not be used, and may be considered as only being installed. Accordingly, it is preferable that the ports 80a and 80b are used for the other purposes and should be used efficiently.
In other words, the network relay apparatus 10 cannot effectively use the ports 80a and 80b that are installed in the data relay unit 60 to achieve the communication path of the spare system.